The power output of many multi-stage steam turbine systems is controlled by sliding pressure control of steam from a steam generator in order to reduce the pressure of steam at the high pressure turbine inlet or steam chest. Steam turbines which utilize this sliding pressure method are often referred to as full arc turbines because all steam inlet nozzle chambers are active at all load conditions. Full arc turbines are usually designed to accept exact steam conditions at a rated load in order to maximize efficiency. By admitting steam through all of the inlet nozzles, the pressure ratio across the inlet stage, e.g., the first control stage, in a full arc turbine remains essentially constant irrespective of the steam inlet pressure. As a result, the mechanical efficiency of power generation across the first control stage may be optimized. However, as power is decreased in a full arc turbine, there is an overall decline in efficiency, i.e., the ideal efficiency of the steam work cycle between the steam generator and the turbine output, because sliding pressure reduces the energy available for performing work. Generally, the overall turbine efficiency, i.e., the actual efficiency is a product of the ideal and the mechanical efficiency of the turbine.
More efficient control of turbine output than is achievable by the sliding pressure method only has been realized by the technique of dividing steam which enters the turbine inlet into isolated and individually controllable arcs of admission. In this method, known as partial-arc admission, the number of active first stage nozzles is varied in response to load changes. Partial arc admission turbines have been favored over full arc turbines because a relatively high ideal efficiency is attainable by sequentially admitting steam through individual nozzle chambers at constant pressure, rather than by sliding pressure over the entire arc of admission. The benefits of this higher ideal efficiency are generally more advantageous than the optimum mechanical efficiency achievable across the control stage of full arc turbine designs. Overall, multi-stage steam turbine systems which use partial-arc admission to vary power output operate with a higher actual efficiency than systems which vary steam pressure across a full arc of admission. However, partial-arc admission systems in the past have been known to have certain disadvantages which limit the efficiency of work output across the control stage. Some of these limitations are due to unavoidable mechanical constraints, such as, for example, an unavoidable amount of windage and turbulence which occurs as rotating blades pass nozzle blade groups which are not admitting steam.
Furthermore, in partial-arc admission systems the pressure drop (and therefore the pressure ratio) across the nozzle blade groups varies as steam is sequentially admitted through a greater number of valve chambers, the largest pressure drop occurring at the minimum valve point (fewest possible number of governor or control valves open) and the smallest pressure drop occurring at full admission. The thermodynamic efficiency, which is inversely proportional to the pressure differential across the control stage, is lowest at the minimum valve point and highest at full admission. Thus, the control stage efficiency for partial-arc turbines decreases when power output drops below the rated load. However, given the variable pressure drops across the nozzles of a partial-arc turbine, it is believed that certain design features commonly found in partial-arc admission systems can be improved upon in order to increase the overall efficiency of a turbine. Because the control stage is an impulse stage wherein most of the pressure drop occurs across the stationary nozzles, a one percent improvement in nozzle efficiency will have four times the effect on control stage efficiency as a one percent improvement in the efficiency of the rotating blades. Turbine designs which provide even modest improvements in the performance of the control stage nozzles will significantly improve the actual efficiency of partial-arc turbines. At their rated loads, even a 0.25 percent increase in the actual efficiency of a partial-arc turbine can result in very large energy savings. Sliding or variable throttle pressure operation of partial-arc turbines within some valve loops also results in improved turbine efficiency and additionally reduces low cycle fatigue. The usual procedure is to initiate sliding pressure operation on a partial-arc admission turbine at flows below the value corresponding to the point where half the control valves are wide open and half are fully closed, i.e., 50% first stage admission on a turbine in which the maximum admission is practically 100%, if sliding pressure is used to the lowest available pressure limit. In comparison, sliding pressure operation is most efficient in a full arc turbine when initiated at maximum steam flow. If sliding pressure is initiated in a partial-arc turbine at a higher flow (larger value of first stage admission), there is a loss in performance. However, in a turbine having multiple valves, sliding from any admission above 50% eliminates a considerable portion of each of the valve loops (valve throttling) which would occur with constant throttle pressure operation. Elimination of such valve loops improves the turbine heat rate and its efficiency.
FIG. 1 illustrates the effect of sliding pressure control in a partial-arc steam turbine having eight control valves. The abscissa represents values of load while the ordinate values are heat rate. Line 14 represents a locus of ideal points for constant pressure with sequential valve control (partial-arc admission) operation. Dotted lines 16, 18, and 20 represent actual valve loops for a finite number of valves. The valve loops result from gradual throttling of each of a sequence of control or governor valves. Sliding pressure operation from 100% admission is indicated by line 22. Note that some of the valve loop 16 is eliminated by sliding pressure along line 22 but that heat rate (the reciprocal of efficiency) increases disproportionately below the transition point 24. Line 26, showing sliding pressure from the 87.5% admission point, provides similar improvement for valve loop 18 down to transition point 28. Similarly, sliding from 75% admission, line 30, improves operation over valve loop 20. Each of these valve loops represents higher heat rates and reduced efficiency from the ideal curve represented by line 14. Each valve point on line 14 represents a condition in which each valve is either fully open or fully closed.
FIGS. 1, 2, and 3 illustrate the operation of an exemplary steam turbine using one prior art control. FIG. 1 shows the locus of full valve points, line 14, with constant pressure operation at 2535 psia. The valve points are identified at 50%, 75%, 87.5% and 100% admission with the valve loops identified by the lines 16, 18, and 20. Sliding pressure is indicated by lines 22, 26 and 30. Starting at 100% admission, about 806MW for the exemplary turbine system, load is initially reduced by keeping all eight control valves wide open and sliding throttle pressure by controlling the steam producing boiler. When the throttle pressure, line 22, reaches the intersection point 24 with the valve loop 16, the throttle pressure is increased to 2535 psia while closing the eighth control valve to an admission value corresponding to point 24. The control valve would continue to close as load is further reduced while maintaining a constant 2535 psia throttle pressure until this valve is completely closed at which point the turbine is operating at 87.5% admission. To further reduce load, valve position is again held constant, seven valves fully open, and throttle pressure is again reduced until the throttle pressure corresponds to the intersection of the sliding pressure line 26 and the valve loop 18 at point 28. To reduce load below point 28, the pressure is increased to 2535 psia and the seventh valve is progressively closed (riding down the valve loop) until it is completely closed. The admission is now 75%. To reduce load still further, the pressure is again reduced with six valves wide open and two fully closed until the sliding pressure line 30 reaches the intersection point 32 with the valve loop 20. Then the operation of raising throttle pressure and closing of a control valve is repeated for any number of valve loops desired. The variation in throttle pressure is illustrated in FIG. 2. The sloped portions 44 of line 46 correspond to sliding pressure operation with constant valve position. The vertical portions 48 correspond to termination of sliding pressure and transition to valve throttling. The horizontal portions 50 correspond to operation at constant steam pressure with control valve throttling such as by riding down the valve loop while reducing load at constant pressure. FIG. 3 shows the improvement in heat rate as a function of load. The line 52 represents the difference between valve loop performance at constant pressure and the performance using sliding pressure between valve points.
The performance improvements shown in FIGS. 1 and 3 are based on the assumption that the boiler feed pump discharge is reduced as the throttle pressure is reduced. If it is not reduced proportionally, the improvement is reduced since the energy required to maintain discharge pressure remains high. In the prior art system, a signal is sent to the feed pump-feed pump drive system to reduce pressure. In reality, however, the feed pump is followed by a pressure regulator in order to eliminate the need for constant adjustment of pump speed and the occurrence of control instability and hunting because of small variations in inlet water pressure to the boiler, resulting from perturbations in flow demand. The regulator, then, does more or less throttling which changes pump discharge pressure and therefore the flow that the pump will deliver. The pump speed is held constant over a desired range of travel of the regulator valve. When the valve travel gets outside these limits, the pump speed is adjusted to move the valve to some desired mean position. As a consequence, the pump discharge pressure does not equal the minimum allowable value (throttle pressure plus system head losses) and so the performance improvement is not as large as shown by FIGS. 1 and 3. In addition, in order to achieve quicker load response, the regulator valve is usually operated with some pressure drop so that if there is a sudden increase in load demand, the valve can open quickly and increase flow. The response of the pump and its drive is slower than the response of the regulator valve.
While the combination of sliding throttle pressure and control valve positioning provides a significant improvement in heat rate, Applicant has found that the optimum transition point for switching from one mode to the other varies from turbine to turbine and over the life of a turbine. In particular, in addition to the factors mentioned above, other parameters such as condenser pressure, reheat temperature, and reheater pressure drop may vary from design values. Such variations cause a shift in the load at which transition occurs. Moreover, because of blading manufacturing tolerances, the transition points (loads) when going from sliding to constant throttle pressure operation differ from those obtained from performance calculations.
U.S. Pat. No. 4,297,848, issued Nov. 3, 1981 and assigned to the assignee of the present invention, attempted to overcome the optimization problem by using impulse chamber pressure to establish the transition point. The procedure described therein required perturbing the boiler pressure and measuring the electrical load. Because of uncertainties in load measurement and complexity of the perturbation, the transition point may occur at a less than optimum value.